Out-of-field doses may affect the formation of secondary cancers, especially in radiosensitive organs, in patients treated with radiotherapy. The aim of this study is to investigate the in-field dose and out-of-field dose accuracy of Eclipse's analytic anisotropic algorithm (AAA) and pencil beam convolution (PBC) algorithms using TLDs. A tissue equivalent phantom containing a total of 21 measurement points at a depth of 5 cm from the anterior and posterior was created. Using Eclipse AAA and PBC algorithms in TPS, 100 MU for AP/PA fields and 95 cm source-skin distance (SSD) were planned. In-field measurement points including isocenter were 3, 5, 7 and 11 points for 3x3, 5x5, 7x7 and 10x10 cm2, respectively. Measuring points outside the field edge were 38, 36, 34 and 30 points for 3x3, 5x5, 7x7 and 10x10 cm2, respectively. In-field point dose values calculated by TPS for different fields were compared with TLD doses measured at the same location. The difference between in-field dose estimation and TLD measurements of both algorithms was generally below 1%. The difference between TPS and TLD was found to be 4.41% for the 10x10 cm2 irradiation field, due to the field edge at a distance of 5 cm from the isocenter. As the field size decreased, the out-of-field dose calculation performance of the AAA and PBC algorithms was adversely affected. For the 10x10 cm2 irradiation field, the TLD measurements and the out-of-field point dose difference of the PBC algorithm were found to be 39.40%. This difference was at most 12.06% for the AAA algorithm. The Eclipse TPS is good at calculating the in-field dose but underestimates the off-field dose. In out-of-field dose calculation, the AAA algorithm gives more accurate results than the PBC algorithm. Additionally, the smaller the field size, the worse the outfield dose accuracy. The use of in vivo dosimeters is recommended in order to estimate the out-of-field dose with great accuracy in radiotherapy.