In this research work, an innovative method for measurement of thermal conductivity of a small volume of liquids, microsphere, and the single cancer cell is demonstrated using a micro-pipette thermal sensor (MPTS). The method is based on laser point heating thermometry (LPHT) and transient heat transfer. When a single pulse of a laser beam heats the sensor tip which is in contact with the surrounding liquids or microsphere/cells, the temperature change in the sensor is reliant on the thermal properties of the surrounding sample. We developed a model for numerical analysis of the temperature change using the finite element method (FEM) in COMSOL. Then we used MATLAB to fit the simulation result with experiment data by multi-parameter fitting technique to determine the thermal conductivity. To verify the accuracy in the measurement of the thermal conductivity by the MPTS method, a 10µl sample of de-ionized (DI) water, 50%, and 70% propylene glycol solution were measured with deviation less than 2% from reported data. Also, to demonstrate that the method can be employed to measure microparticles and a single spherical cell, we measured the thermal conductivity of poly-ethylene microspheres with a deviation of less than 1% from published data. We estimated the thermal conductivity of two types of cell culture growth media for the first time and determined the thermal conductivity of cancerous Jurkat Clone E6-1 to be 0.538 W/m.K ± 2%. Using the sensor of 1-2μm tip size, we demonstrated the MPTS technique as a highly accurate technique for determining the thermal conductivity of microfluidic samples, microparticles, biological fluids, and a non-invasive method for measuring the thermal conductivity of single cancer cell. This MPTS technique can be beneficial in developing a diagnosis method for the detection of cancer at an early stage. We also compared three effective thermal conductivity models for determining the weight percentage of Jurkat cell, considering water and protein as the major constituents. We discovered that a combination of Maxwell-Euken and effective medium theory model provides the closest approximation to published data and, therefore, recommend for the prediction of the cell composition.