Survival and reproduction of a linear population of Bufo punctatus was studied during 1964 in an often waterless, mile—long section of a canyon of the Colorado Desert of California. Reproduction had been unsuccessful in earlier years of record because of little or no springflow and infrequency of summer flashfloods. In 1964, springflow began on March 23 and retreated rapidly in early May. There was one summer flashflood. Toads did not emerge from underground winter retreats until evening air temperature was around 18°C and did not breed until the weather was some 6 degrees warmer. The relatively few females who oviposited, mainly during April 9 to 6, produced and estimated 34,000 eggs. The lowest downcanyon site of oviposition in the spring was 5,370 ft above the lowest downcanyon site in 1962. Home range up and down the canyon during the breeding season was about 600 ft at the most for males and 1,500 ft more for females. Some trilling males went underground when the stream dried at their sites; others migrated upcreek as much as 2,000 ft to new trilling sites beside water. Many females did not migrate, while others showed a tendency to wander great distances (2,700 ft maximum record) towards or away from water. When the stream dried completely all toads disappeared but emerged after the one flashflood of summer in order to hydrate. Tracks then occurred 1,300 ft farther downcreek than any individual had been seen in spring. Very limited breeding was accomplished by individuals who were different from the spring breeders. Rate of larval development in spring (approximately 2 months to metamorphosis) was retarded by coolness of the water. There were four different stages of larval behavior: 1) egg—site aggregations, when tadpoles massed at sites of oviposition; 2) deepwater aggregations, when the mass moved to deepest available water; 3) feeding dispersal when tadpoles dispersed in search of food; and 4) thermal aggregations, when they congregated in warmest water at margins of pools by day and at bottoms by night. Nearly all tadpoles were killed by desiccation both in spring and summer. Since some toe—marked adults were captured after 4 years, longevity may be a reason for survival of populations in severe habitats where successful reproduction is rare.