In the framework of microgenerators, we present in this paper the key points for energy harvesting from temperature using ferroelectric materials. Thermoelectric devices profit from temperature spatial gradients, whereas ferroelectric materials require temporal fluctuation of temperature, thus leading to different applications targets. Ferroelectric materials may harvest perfectly the available thermal energy whatever the materials properties (limited by Carnot conversion efficiency) whereas thermoelectric material's efficiency is limited by materials properties (ZT figure of merit). However, it is shown that the necessary electric fields for Carnot cycles are far beyond the breakdown limit of bulk ferroelectric materials. Thin films may be an excellent solution for rising up to ultra-high electric fields and outstanding efficiency. Different thermodynamic cycles are presented in the paper: principles, advantages, and drawbacks. Using the Carnot cycle, the harvested energy would be independent of materials properties. However, using more realistic cycles, the energy conversion effectiveness remains dependent on the materials properties as discussed in the paper. A particular coupling factor is defined to quantify and check the effectiveness of pyroelectric energy harvesting. It is defined similarly to an electromechanical coupling factor as k2=p2theta0/(epsilontheta33cE), where p, theta0, epsilontheta33, cE are pyroelectric coefficient, maximum working temperature, dielectric permittivity, and specific heat, respectively. The importance of the electrothermal coupling factor is shown and discussed as an energy harvesting figure of merit. It gives the effectiveness of all techniques of energy harvesting (except the Carnot cycle). It is finally shown that we could reach very high efficiency using 1110.75Pb(Mg1/3Nb2/3)-0.25PbTiO3 single crystals and synchronized switch harvesting on inductor (almost 50% of Carnot efficiency). Finally, practical implementation key points of pyroelectric energy harvesting are presented showing that the different thermodynamic cycles are feasible and potentially effective, even compared to thermoelectric devices.