Objective To estimate the completeness of malaria notification to the public healthcare system (PHCS) and to describe retrospectively data of malaria cases in Tunisia. Methods We conducted a retrospective epidemiological survey using a standardized questionnaire for all cases of malaria reported to PHCS and those diagnosed in parasitological laboratory or infectious disease service between January 2002 and December 2007. To estimate the total number of cases, we used a two sources capture-recapture analysis. Results After record-linkage and cross-validation 317 cases of malaria were identified, of whom 231 were notified, resulting in an observed under-notification of 17%. The estimated number of malaria cases using capture-recapture analysis was 366.3 (95% CI: 335.8-396.8) for the period of study with a completeness of 63.1% which increased from 44.8% for 2002 to 78.7% for 2007. One hundred and sixty two patients (51.1%) had been born in sub-Saharan Africa, 113 (35.6%) in Tunisia, 35 (11.0%) in North Africa and 7 (2.2%) in Europe with predominance of men (87.1% of all cases). The median age was 25.0 years (21-30) for sub-Saharan Africans, 38.0 years (23.5-45.5) for North Africans, 38.5 years (30.75-38.5) for Tunisian and 39.0 (26-43) for European ( P<0.01). The most predominant malaria species was Plasmodium falciparum with 216 cases (72.5%), and the most frequent area of acquisition was sub-Saharan Africa. In our study, information on compliance with malaria prophylaxis was only sporadically available and 34% of infected individuals had not used any chemoprophylaxis. Our study showed delayed identification of malaria that indicated a deficit in medical awareness and management of this infection. Conclusions Our survey has marked variety in the type and availability of key data and shown an underreporting of malaria cases. Furthermore, it demonstrates that the two different sources of malaria registration are substantially incomplete. Of particular interest is the observation that a considerable number of patients could only be found in the records of PHCS, they were unknown to the laboratories, although malaria confirmation by thick or thin smear is obligatory in Tunisia.