This paper argues against the current view that the apsidal form was Buddhist in origin and that apsidal Hindu temples are essentially Buddhist shrines subsequently converted to Hinduism. It also counters the linear view of religious change, which suggests that the Hindu temple came into its own after the decline of Buddhism in the fourth-fifth centuries ad. Instead, the paper establishes that the apsidal form was part of a common architectural vocabulary widely used from the second century bc onwards not only for the Buddhist shrine, but also for the Hindu temple and several local and regional cults. The historical development of the apsidal shrine is traced at three levels. One is at the macro-level of the Indian subcontinent from the second century bc to the nineteenth century ad. Second is the location of the apsidal form in the religious landscape of a particular site, viz. that of Nagarjunakonda in the lower Krishna valley in Andhra. Finally, the paper highlights the archaeological excavations conducted at the apsidal temple of Gudimallam also in Andhra to trace the history of the religious site from the second century bc to the twelfth century ad, the objective being to highlight the changing nature of the Hindu temple. The paper thus makes a case for plurality of religious beliefs and practices in ancient South Asia as against the prevailing view that these local and regional cults were gradually subsumed under the mantle of Sanskritization starting from the fourth-fifth centuries onwards.