When determining requirements for a SONAR system as a whole, typically very high-level requirements such as number of detections and total false alarms allowed per unit time must be converted into specific requirements for system components. These requirements may include array requirements like hydrophone count and position. Validation of probability of detection via sea test is fraught with complications, not least of which is environmental uncertainty. As a precursor, it is generally advisable to approach and attempt to validate models for each term in the SONAR equation one at a time. The approaches to validation vary depending on the term and can range in complexity from comparison to analytical solutions, to comparison with alternate available models, up through data collections at sea. Each of these approaches comes with its own challenges and limitations, which are discussed.