The cost of high-level quantum chemical calculations is known to increase sharply with the system size, restricting their applicability to small- or medium-sized systems. For treating large molecules with correlated methods, several cost-effective strategies such as fragmentation-based approaches have been developed since 1970. I will discuss the development of molecular tailoring approach (MTA) in my laboratory [1] during the last two decades. It began with the aim of accurate calculation [2, 3] of molecular electron density and electrostatic potential of large molecules in 1994. Geometry optimization of large molecules, based on the set inclusion-exclusion principle, was built within the MTA framework around 2006 [4] and applied extensively to several large molecules and molecular clusters [5, 6]. MTA enables correlated calculations on off-the-shelf hardware, although it leads to errors of a few millihartrees (mH) vis-à-vis the respective full calculation energies. A grafting methodology, incorporating correction for the missing interactions due to fragmentation, was introduced and applied successfully for accurate energy estimation of (H2O)20 clusters [6]. Furthermore, the MTA, in conjunction with grafting correction, was effectively applied for an accurate yet inexpensive calculation of vibrational IR and Raman spectra of molecular clusters and real proteins [7, 8]. An automated code for fragmentation, balancing accuracy and efficiency is recently developed and available to users [9]. The talk will demonstrate the effective parallelization of the MTA code, pointing to the enormous potential of MTA for studying energetics and spectra of large molecules and clusters on a limited hardware.