PURPOSE: The aim of this study was to develop a method to derive the midplane dose [i.e., the two-dimensional (2D) dose distribution in the middle of a patient irradiated with high-energy photon beams] from transmission dose data measured with an electronic portal imaging device (EPID). A prerequisite for this method was that it could be used without additional patient information (i.e., independent of a treatment-planning system). Second, we compared the new method with several existing (conventional) methods that derive the midline dose from entrance and exit dose measurements. METHODS AND MATERIALS: The proposed method first calculates the 2D contribution of the primary and scattered dose component at the exit side of the patient or phantom from the measured transmission dose. Then, a correction is applied for the difference in contribution for both dose components between exit side and midplane, yielding the midplane dose. To test the method, we performed EPID transmission dose measurements and entrance, midplane, and exit dose measurements using an ionization chamber in homogeneous and symmetrical inhomogeneous phantoms. The various methods to derive the midplane dose were also tested for asymmetrical inhomogeneous phantoms applying two opposing fields. A number of combinations of inhomogeneities (air, cork, and aluminum), phantom thicknesses, field sizes, and a few irregularly shaped fields were investigated, while each experiment was performed in 4-, 8-, and 18-MV open and wedged beams. RESULTS: Our new method can be used to assess the midplane dose for most clinical situations within 2% relative to ionization chamber measurements. Similar results were found with other methods. In the presence of large asymmetrical inhomogeneities (e.g., lungs), discrepancies of about 8% have been found (for small field sizes) using our transmission dose method, owing to the absence of lateral electron equilibrium. Applying the other methods, differences between predicted and measured midplane doses were even larger, up to 10%. For large field sizes, the agreement between measured and predicted midplane dose was within 3% using our transmission dose method. CONCLUSIONS: Using our new method, midplane doses were estimated with a similar or higher accuracy compared with existing conventional methods for in vivo dosimetry. The advantage of our new method is that the midplane dose can be determined in the entire (2D) field. With our method, portal in vivo dosimetry is an accurate alternative for conventional in vivo dosimetry.