We compared Brazilian oral cleft (OC) frequencies between the population-based Brazilian System of Live Birth (SINASC) and the hospital-based Latin American Collaborative Study of Congenital Malformations (ECLAMC), trying to understand the paucity of cleft of lip and palate (CLP) in the first system. SINASC uses the International Classification of Disease version 10 (ICD-10) for congenital defects coding, ECLAMC uses ICD-8 with modifications. In SINASC, the CLP frequency was 1.7 per 10,000 (95% confidence limits 1.7-1.8), cleft lip (CL) 1.6 (1.5-1.7), and cleft palate (CP) 2.0 (1.9-2.1). In ECLAMC, the CLP frequency was 10.4 per 10,000 (9.0-12.1), CL 5.5 (4.5-6.7), and CP 4.4. (4.5-6.7). In SINASC, only 33% of the oral clefts were CLP, versus 51% in ECLAMC. Part of this discrepancy may have been due to the relative excess of CP and CL cases. Although congenital defect frequencies are usually lower in population than in hospital-based registries, differences in the proportion of the main OC categories are not expected and are probably due to ICD-10 coding issues, such as lumping of unilateral CL and CL without other specifications. ICD-10 codes, whose deficiency for oral clefts is fully explained in the literature, lack modifiers for severity, or clinical subtypes. This paper shows the practical aspect of the ICD-10 system deficiency in capturing cleft lip and palate (CLP) subtypes, as demonstrated in SINASC covering three million births per year. Such errors are expected to occur in any registry that uses the ICD-10 coding system, and must be adjusted, given its relevance worldwide.