Acute flaccid paralysis (AFP) surveillance data from India were analysed to examine sensitivity of poliovirus isolation from stool specimens and the added sensitivity obtained from collection of a second stool specimen. Analysis was restricted to Indian AFP cases, 1998–2000, with two adequate stool specimens. The proportion of cases confirmed with wild poliovirus isolation by the second specimen only was calculated, regardless of specimen quality. Overall specimen sensitivity (1998–2000) was 81% using the first specimen, 78% using the second, and 96% using both. Sensitivity increased from 1998 to 2000, with slightly higher sensitivity each year for the first specimen. The second specimen increased sensitivity by 15% overall and contributed more when the first specimen was collected late or was in poor condition. As wild poliovirus disappears, increased sensitivity provided by a second stool specimen may reduce the risk of missing circulating virus.