Magnetic force microscopy (MFM) is a very powerful technique, which can potentially be used to detect and localize the magnetic fields arising from nanoscopic magnetic domains, such as magnetic nanoparticles. However, in order to achieve this, we must be able to use MFM to discriminate between magnetic forces arising from the magnetic nanoparticles and nonmagnetic forces from other particles and sample features. Unfortunately, MFM can show a significant response even for nonmagnetic nanoparticles, giving rise to potentially misleading results. The literature to date lacks evidence for MFM detection of magnetic nanoparticles with nonmagnetic nanoparticles as a control. In this work, we studied magnetite particles of two sizes and with a silica shell, and compared them to nonmagnetic metallic and silica nanoparticles. We found that even on conducting, grounded substrates, significant electrostatic interaction between atomic force microscopy probes and nanoparticles can be detected, causing nonmagnetic signals that might be mistaken for a true MFM response. Nevertheless, we show that MFM can be used to discriminate between magnetic and nonmagnetic nanoparticles by using an electromagnetic shielding technique or by analysis of the phase shift data. On the basis of our experimental evidence we propose a methodology that enables MFM to be reliably used to study unknown samples containing magnetic nanoparticles, and correctly interpret the data obtained.