Summary. The paper describes the dynamics of trends in the number of specialists and the differences that have arisen over a 14-year period among individual specialties. Shown and analysed separately is the basic demographic information (obtained from the National Health Workers Registry) about the specialists working in Croatia in 2000. The data were used to illustrate the two options for planning them: by statistical method for analysing a time sequence and by prediction of specialist needs by anticipated old age retirement. At the end of 2000 there were 7,383 specialists, or 65.8% of the physician total employed in Croatian health service. Of these, 3,664 (49.6%) were females, with the median for females and males being 47 and 50 years, respectively. In relation to 1986, there was a twofold increase or greater in the number of maxillofacial surgeons (basic index 260; average rate of annual increase 7.1%), followed by psychiatrists, epidemiologists, child surgeons and medical cytologists. The paper uses the Holt two-parameter method of exponential smoothing of the time sequence made up of the number of medical specialists per 100,000 population in 1980–99. Assuming maintained average annual growth of three specialists, this prognostic procedure predicted specialist density of 185/100,000 population for 2009. The feasibility of a more direct planning is illustrated with individual specialist sex and age data at certain age of life. It permits estimating the number of specialists to leave the health system for the most common reason, i.e. old-age retirement. On the set criterion of lady doctors reaching the age of 60 and their male colleagues 65 years, we predicted for each medical speciality the annual number of specialists due to retire for old age by the year 2020. In some specialties, the difference in the number of specialists anticipated to go into old-age retirement in two successive years will reach close to 50% or even slightly higher. In this way, the necessary number of replacement specialists for the retired can be estimated for each medical specialty by taking note of the critical years when, because of great variation in age composition, a larger number of specialists could simultaneously leave their posts. For these critical years, a sufficient number of specialists should be ensured by harmonizing our length of education and training curricula with those in force in the EU countries. The number of specialists and their composition will also be changing, depending on the health system’s strategic solution to the need for specialist work in the Family Medicine Service.