Current U.S. Environmental Protection Agency regulatory guidelines for developmental neurotoxicity emphasize functional categories such as motor activity, auditory startle, and learning and memory. A single test of some simple form of learning and memory is accepted to meet the latter category. The rationale for this emphasis has been that sensitive and reliable methods for assessing complex learning and memory are either not available or are too burdensome, and that insufficient data exist to endorse one approach over another. There has been little discussion of the fact that learning and memory is not a single identifiable functional category and no single test can assess all types of learning and memory. Three methods for assessing complex learning and memory are presented that assess two different types of learning and memory, are relatively efficient to conduct, and are sensitive to several known neurobehavioral teratogens. The tests are a 9-unit multiple-T swimming maze, and the Morris and Barnes mazes. The first of these assesses sequential learning, while the latter two assess spatial learning. A description of each test is provided, along with procedures for their use, and data exemplifying effects obtained using developmental exposure to phenytoin, methamphetamine, and MDMA. It is argued that multiple tests of learning and memory are required to ascertain cognitive deficits; something no single method can accomplish. Methods for acoustic startle are also presented.