The actions are sub-divided into eleven broad categories as given below.
1. Modification of regime; 2. Land transformation and construction;
3. Resource extraction; 4. Processing;
5. Land alteration; 6. Resource renewal;
7. Changes in traffic; 8. Waste emplacement and treatment;
9. Chemical treatment; 10. Accidents
11. Other (unspecified). .

In a similar way the environmental items are classified into five categories and eleven sub-categories:
1. Physical and chemical characteristics;
a. earth,
b. water,
c. atmosphere,
d. processes
2. Biological conditions;
3. Ecological relationships;
4. Others (unspecified).
The complete list of 'actions' and environmental items included in the Leopold matrix is reproduced in many EIA publications, including Canter (1977), Warner (1973) and Westman (1985).

The items included in the Leopold matrix are not above criticism. For example 'modification of regime, which include 'alteration of ground-water hydrology', 'weather modification' etc., has been included as actions. Logically these are consequences of actions and not actions per-se. The action 'blasting and drilling' is double counted, first under the category 'land transformation and construction' and then under 'resource extraction'. 'Blasting and drilling', 'surface excavation', 'mineral processing', 'mine sealing and waste control', 'strip mining rehabilitation', 'reforestation', 'emplacement of tailings, spoils and overburden', etc. are included as actions, but 'mining and quarrying' finds a mention under environmental items, that too under the category of cultural factors. It is not clear how 'accidents' can be included under actions! 'Waste recycling' is included as an environmental item. Many of the actions are mutually exclusive and therefore many of the cells need not be filled up at all. Thus on one side the matrix is confusing and on the other many of the entries are redundant.

Leopold matrix can be used to identify, measure and interpret impacts. Because of mutual exclusivity of many items, for any project only some of the 100 actions and 88 environmental items are likely to be involved. Impacts are identified by following a two step process:
1. Identification of all the actions that are part of the proposed project; and,
2. Under each of the proposed actions, placing a slash in each cell where interaction with the environmental item is possible. The slash is placed from upper right to lower left corner of the cell.

The interaction is then described in terms of 'magnitude' and 'importance'. Leopold matrix incorporates guidelines for characterisation of impacts in terms of magnitude or importance on a common 1-10 scale. Impacts are measured by placing a numeric value (between 1 and 10) above the slash to indicate the 'magnitude' of the possible impact. 10 represents a large magnitude and a small magnitude is represented by 1. Unsigned values indicate adverse impact while a plus sign is placed to indicate a beneficial impact. EIA practitioners, however, often prefer to place a minus sign to indicate a negative impact. Impact magnitude is assigned by the investigator on the basis of an objective or quasi-objective evaluation of facts related to the anticipated impact. Impact magnitude is recorded in the upper half of the interaction cell.

Impacts are interpreted in terms of significance or importance on a scale of 1-10 and the importance value is recorded in the lower half of the interaction cell. Importance is used to analyze the consequences of a proposed action. The value of 10 represents a very significant interaction and value of 1 indicates an interaction of relatively low importance. Numerical values for importance are assigned by the investigator based on subjective judgement.

Leopold et al. (1971) emphasised that the decision logic and the thought process involved in assigning both the magnitude and importance score should be clearly recorded in the accompanying text.

As may be observed Leopold et al. (1971) used the term 'magnitude' in the sense of degree, extensiveness or scale. Though magnitude rating may be expected to be based on objective measurement using physical units wherever possible, the conversion to the magnitude scale of 1 to 10 is essentially subjective. No guideline or rule is provided for converting the measured values to magnitude scale. The magnitude scale used in Leopold Matrix should not, therefore, be considered as a scale used for commensuration. In fact Leopold et al. (1971) themselves cautioned that 'no two boxes on any one matrix are equatable'. However use of common numerical scale only adds to the confusion and many EIA scholars, including Lohani and Thanh (1977), were tempted to arrive at a summarised project impact score by assuming that the magnitude and impact values are expressed using a commensurate scale. Such attempts are fundamentally flawed because a rating in Leopold matrix relates only on the specific environmental attribute under consideration and does not apply to other parameters. The Leopold matrix may be used as an effective impact identification tool. That will, however, require careful modifications of the listed parameters and activities. Use of subjective scaling to express impact significance permits taking into account special conditions prevailing at a site or to account for the project specific situations. In its original form the Leopold matrix does not include many important parameters. However, the Leopold matrix can be easily modified to suit specific situations. An important drawback of Leopold matrix is that it fails to differentiate between direct and indirect impacts. Non project impacts cannot be identified by this methodology. Because of the inherent subjectiveness the Leopold matrix leaves ample scope for individual bias to affect the assessment. The matrix therefore, lacks replicability. As has been mentioned before the entries under actions and environmental characteristics are difficult to comprehend on a logical basis.

In addition Warner (1973) contended the validity of the following entries which are difficult to distinguish in terms of their separate effects on environmental characteristics: "alteration of drainage", "river control and flow modification", and "canals". Because of these ambiguities, subjectivity, and lack of guidelines for assigning importance and magnitude values, the Leopold approach when applied gives very low replicability. The structure of the matrix is flexible and so is the resource requirement. The subjective scaling of impact magnitude can be equally applicable to a very precise estimate and an intuitive assessment. A minimum requirement is the description of the pre-project environment and the description of important project features. Neither of these are very demanding and each of them can be accomplished with low resource requirement albeit by compromising the reliability and exactitude. Leopold approach does not provide any guideline regarding adequacy of data. An EIA analyst is therefore free to arrive at the best estimate based on the data available.

Leopold matrix may at times require a very large number of impacts to be analysed. A maximum number of 8800 impact cell may have to be analysed which would require 17600 values of impact magnitudes and importance to be filled. This is an 'intolerable situation' (Warner 1973). Two large matrices prepared for two alternatives are difficult to compare. A large matrix with many completed cells is difficult to interpret in terms of overall result (Westman 1985). Time, manpower, cost and technology requirements for Leopold matrix are also flexible. It is, however, obvious that involvement of more number of analysts would be helpful in reducing individual bias. Similarly a little extra time may have to be provided to analyse indirect effects. For an analysis of impacts in Leopold format 'cost involved' is essentially a function of time and data adequacy. Leopold matrix is easily amenable to various degrees of technology - from sophisticated computer modelling to intuitive estimates.

To sum up the author infers that Leopold type matrices may be used for impact identification. Effectiveness of Leopold matrix is severely limited in terms of measurement and interpretation of impacts. It does not have any summarisation scheme and impact communication to lay people become difficult. However, it displays cause-effect relationship and thereby assists decision-makers in monitoring and mitigation planning. Because of the flexibility of resource requirements this methodology can be applied in developing countries. In order to be applicable to mining projects the entries under both project actions and environmental attributes will have to be modified to a significant extent. If applied with caution Leopold type matrices may be expected to satisfactorily identify the 'impact-causing factors' and the environmental attributes affected by such factors when the project comes into life. Leopold matrix is useful also for characterising direct impacts, but like most other checklists and matrices it presents a fragmented view of environmental systems. This is the most important factor inhibiting its possible upgradation as a tool to evaluate environmental sustainability of a proposed project. That is because environmental sustainability requires, as a minimum, a holistic view on the environment.

This methodology cannot be used to determine indirect impacts (second or higher order impacts).