Background: The reliability of information sources, as well as the precision of coverage indicators are open questions for immunisation coverage surveys. Methods: A sample survey of 323 children from two birth cohorts (1991–1992 and 1979–1981) was done. The sample was selected from seven health districts out of 51 in the Lazio region. The information was collected from vaccination cards or from parents, using a mail-in questionnaire. We computed three indicators of coverage at 24 months of age for polio, diphtheria–tetanus, pertussis (P), hepatitis B, and measles (M), depending on the value given to parental recall. We also compared up-to-date (UTD) (immunised children at 24 months) and age-appropriate (months at risk per child) indicators at the local level. Results: Response rate was 68.2%. The two birth cohorts were similarly immunised for polio (coverage 95%); the younger cohort had higher coverage levels for diphtheria–tetanus (93% vs. 75%), for P (30% vs. 22%), for M (43% vs. 16%) than the older one, even on the basis of parental recall. In the older cohort, diphtheria–tetanus, P and M coverage levels were 75, 22 and 15%, respectively, on the basis of parental recall and 66, 17 and 4% on certification. The age-appropriate indicator revealed four districts that completed vaccinations significantly too late, while the UTD indicator showed only one district significantly below the threshold of 95% of immunized children at 24 months. Conclusions: Parental recall and data from cards are comparable sources, but the former is more sensitive than the latter in retrospective studies. Verbal recall should be accepted as reasonably reliable in the absence of cards. Age-appropriate indicators give a more complete evaluation of the susceptibility in the population, and are more precise at the local level.